Leo Mckee-Reid

RIDArena (Rogue Internal Deployment Arena) is an AI control evaluation environment that simulates an AI lab’s R&D infrastructure. Researchers can use it to study how an agent can covertly sabotage the lab’s monitoring system and deploy another rogue agent without oversight. RIDArena can support sabotage, control, and monitor evaluations in...

Hello! I'm part of the AI club at the University of Victoria. We want to help expose students and professors at our university to the ideas of AI alignment and AI ethics, and so we are putting together an online conference this November 25th. Our club website Eventbrite page We...